This isn't quite ready yet. Version 3.3 will allow imports, transforms of items into ez objects.
To do this you would need to write some php code that read your .csv and then created ez objects. Not too hard but its a steep learning curve to start with.
I apreciate the quick response. Tell me though, is it possible to do other types of imports? We have a lot of data in oracle tables that we were displaying with CFM pages, but really the data is pretty cut and dried HTML. Can we go directly to the MySQL database in put this in directly as an article type for instance or is their any type of import script for this?
If you need to keep your oracle RDBMS, you may as well create your own datatype for the records in the database. When creating objects with this new datatype, you can for instance specify the key to use (which is stored inside ezp tables). This makes it possible that "foreign" data is integrated with ez publish, including all the benefits ...
Not so easy, but may be worth the effort depending on what you want to do.
We face the same problems on the taxonomy ... and finally did it by hand in the admin interface except the few things we could migrate from an ezpublish 2.2 site.
If you have your taxonomy as a folder structure in a file system, you could do this with teh upcoming webdav interface too (perhaps create the folders first with a simple script if the taxonomy is inside a database or csv file)
I greatly appreciate the response. We don't need to keep the Oracle RDBMS. Was just again wondering if I can find a quick way to import lots of records into Ez PUblish. I'm doing part of the taxonomy by hand via the interface now as a test. Is it possible to just swallow the tables without too much trouble into MySQL and have the data display properly? I haven't looked at the table structure yet.
Thanks again for answering these newbie questions!
I think you need to think about your end aim here and work out the necessary steps.
For example, take a look at creating an ez class and determine whether your data can be modelled in the way you need it to be. If you can create a class with attributes that matches the structure of your data then all you need to do is create the ez objects.
This is, as you might guess, a little more difficult to do. You mention csv data so i assume you have a way of exporting your oracle fields into a handle text file.
Paul mentioned using webdav, which is new in 3.2. You could use that to 'drop' in files, which will create the correct objects (for the files mime type) in the database. For that you may need to alter the webdav code a little to match the type for your class.
However a little script to read your csv file, and create the objects inside ez looks like the best way to go. I've worked on similar things, its just a matter of finding what function calls to use.
Paul
You must be logged in to post messages in this topic!
